Exploratory Analysis

library(tidyverse)
library(patchwork)
library(knitr)
library(dplyr)
library(gganimate)
library(gifski)
library(png)
library(plotly)

demo=foreign::read.xport("nhanes/P_DEMO.XPT") %>%
  janitor::clean_names() %>%
  select(seqn,riagendr,ridageyr,ridreth3,dmdeduc2,indfmpir) %>%
  drop_na() %>%
  rename(gender=riagendr,
         age=ridageyr,
         race=ridreth3,
         education_level=dmdeduc2,
         income_poverty_ratio=indfmpir) %>%
  mutate(gender=case_when(gender==1 ~ "male",
                          gender==2 ~ "female")) %>%
  mutate(race=case_when(race==1 ~ "Mexican American",
                        race==2 ~ "Other Hispanic",
                        race==3 ~ "Non-Hispanic White",
                        race==4 ~ "Non-Hispanic Black",
                        race==6 ~ "Non-Hispanic Asian",
                        race==7 ~ "Other Race")) %>%
  mutate(education_level=case_when(education_level==1 ~ "less than 9th grade",
                                   education_level==2 ~ "9-11th grade",
                                   education_level==3 ~ "high school graduate",
                                   education_level==4 ~ "college or AA degree",
                                   education_level==5 ~ "college graduate or above",
                                   education_level==7 ~ "refused",
                                   education_level==9 ~ "don't know"))


sleep_df=foreign::read.xport("nhanes/P_SLQ.XPT") %>%
  janitor::clean_names() %>% 
  select(seqn,slq300,slq310,sld012,slq320,slq330,sld013,slq050,slq120) %>%
  rename(weekday_slp_time=slq300,
         weekday_wake_time=slq310,
         weekday_slp_hr=sld012,
         weekend_slp_time=slq320,
         weekend_wake_time=slq330,
         weekend_slp_hr=sld013,
         trouble_slp=slq050,
         sleepy_freq=slq120
         )              %>%
  filter(weekday_slp_time!=77777 & weekday_slp_time!= 99999) %>%
  filter(weekday_wake_time!=77777 & weekday_wake_time!= 99999) %>%
  filter(weekend_slp_time!=77777 & weekend_slp_time != 99999) %>%
  filter(weekend_wake_time!= 77777 & weekend_wake_time != 99999) %>%
  filter(trouble_slp!=7 & trouble_slp!= 9) %>%
  filter(sleepy_freq!= 7 & sleepy_freq!=9) %>%
  drop_na() %>%
  mutate(trouble_slp=case_when(trouble_slp==1 ~ "yes",
                               trouble_slp==2 ~ "no")) %>%
  mutate(sleepy_freq=case_when(sleepy_freq==0 ~ "never",
                               sleepy_freq==1 ~ "once a month",
                               sleepy_freq==2 ~ "2-4 per month",
                               sleepy_freq==3 ~ "5-15 per month",
                               sleepy_freq==4 ~ "16-30 per month",
                               ))

slp_df=inner_join(demo,sleep_df,by="seqn") %>%
  na_if("") %>%
  na.omit()
write_csv(slp_df, "data/slp_df.csv")

year_df=read_csv("data/AdultTrends.csv") %>%
  janitor::clean_names() %>%
  pivot_longer(cols=c("female", "male"),
               names_to = "sex",
               values_to = "age_adjust_prev") %>%
  ggplot(aes(x=year,y=age_adjust_prev,col=sex))+geom_line()+geom_point()
year_df+transition_reveal(year)

 edu_plot=slp_df %>%
  filter(weekday_slp_hr<6)%>%
  group_by(education_level,gender) %>%
  summarize(obs=n()) %>% 
  ungroup() %>%
  mutate(education_level=fct_reorder(education_level,obs)) %>%
  ggplot(aes(x=education_level,y=obs,fill=gender))+ geom_bar(stat="identity")+
  viridis::scale_fill_viridis(
    name = "gender",
    discrete = TRUE
  ) + 
  theme(axis.text.x = element_text(angle = -90, vjust = 0.5, hjust=1))

ggplotly(edu_plot)